File: step_down_spec.rb

package info (click to toggle)
ruby-mongo 2.21.3-1
  • links: PTS, VCS
  • area: main
  • in suites: sid
  • size: 14,764 kB
  • sloc: ruby: 108,806; makefile: 5; sh: 2
file content (225 lines) | stat: -rw-r--r-- 6,649 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
# frozen_string_literal: true
# rubocop:todo all

require 'spec_helper'

describe 'Step down behavior' do
  require_topology :replica_set

  # This setup reduces the runtime of the test and makes execution more
  # reliable. The spec as written requests a simple brute force step down,
  # but this causes intermittent failures.
  before(:all) do
    # These before/after blocks are run even if the tests themselves are
    # skipped due to server version not being appropriate
    ClientRegistry.instance.close_all_clients
    if ClusterConfig.instance.fcv_ish >= '4.2' && ClusterConfig.instance.topology == :replica_set
      # It seems that a short election timeout can cause unintended elections,
      # which makes the server close connections which causes the driver to
      # reconnect which then fails the step down test.
      # The election timeout here is greater than the catch up period and
      # step down timeout specified in cluster tools.
      ClusterTools.instance.set_election_timeout(5)
      ClusterTools.instance.set_election_handoff(false)
    end
  end

  after(:all) do
    if ClusterConfig.instance.fcv_ish >= '4.2' && ClusterConfig.instance.topology == :replica_set
      ClusterTools.instance.set_election_timeout(10)
      ClusterTools.instance.set_election_handoff(true)
      ClusterTools.instance.reset_priorities
    end
  end

  let(:subscriber) { Mrss::EventSubscriber.new }

  let(:test_client) do
    authorized_client_without_any_retries.with(server_selection_timeout: 20).tap do |client|
      client.subscribe(Mongo::Monitoring::CONNECTION_POOL, subscriber)
    end
  end

  let(:collection) { test_client['step-down'].with(write: write_concern) }

  let(:admin_support_client) do
    ClientRegistry.instance.global_client('root_authorized').use('admin')
  end

  describe 'getMore iteration' do
    min_server_fcv '4.2'
    require_no_linting

    let(:subscribed_client) do
      test_client.tap do |client|
        client.subscribe(Mongo::Monitoring::COMMAND, subscriber)
        client.subscribe(Mongo::Monitoring::CONNECTION_POOL, subscriber)
      end
    end

    let(:collection) { subscribed_client['step-down'] }

    before do
      collection.insert_many([{test: 1}] * 100)
    end

    let(:view) { collection.find({test: 1}, batch_size: 10) }
    let(:enum) { view.to_enum }

    it 'continues through step down' do
      server = subscribed_client.cluster.next_primary
      server.pool_internal.do_clear
      server.pool_internal.ready
      subscriber.clear_events!

      # get the first item
      item = enum.next
      expect(item['test']).to eq(1)

      connection_created_events = subscriber.published_events.select do |event|
        event.is_a?(Mongo::Monitoring::Event::Cmap::ConnectionCreated)
      end
      expect(connection_created_events).not_to be_empty

      current_primary = subscribed_client.cluster.next_primary
      ClusterTools.instance.change_primary

      subscriber.clear_events!

      # exhaust the batch
      9.times do
        enum.next
      end

      # this should issue a getMore
      item = enum.next
      expect(item['test']).to eq(1)

      get_more_events = subscriber.started_events.select do |event|
        event.command['getMore']
      end

      expect(get_more_events.length).to eq(1)

      # getMore should have been sent on the same connection as find
      connection_created_events = subscriber.published_events.select do |event|
        event.is_a?(Mongo::Monitoring::Event::Cmap::ConnectionCreated)
      end
      expect(connection_created_events).to be_empty
    end

    after do
      # The tests normally operate with a low server selection timeout,
      # but since this test caused a cluster election we may need to wait
      # longer for the cluster to reestablish itself.
      # To keep all other tests' timeouts low, wait for primary to be
      # elected at the end of this test
      test_client.cluster.servers.each do |server|
        server.unknown!
      end
      test_client.cluster.next_primary

      # Since we are changing which server is primary, close all clients
      # to prevent subsequent tests setting fail points on servers which
      # are not primary
      ClientRegistry.instance.close_all_clients
    end
  end

  describe 'writes on connections' do
    min_server_fcv '4.0'

    let(:server) do
      client = test_client.with(app_name: rand)
      client['test'].insert_one(test: 1)
      client.cluster.next_primary
    end

    let(:fail_point) do
      {
        configureFailPoint: 'failCommand',
        data: {
          failCommands: ['insert'],
          errorCode: fail_point_code,
        },
        mode: { times: 1 }
      }
    end

    before do
      collection.find
      admin_support_client.command(fail_point)
    end

    after do
      admin_support_client.command(configureFailPoint: 'failCommand', mode: 'off')
    end

    describe 'not master - 4.2' do
      min_server_fcv '4.2'

      let(:write_concern) { {:w => 1} }

      # not master
      let(:fail_point_code) { 10107 }

      it 'keeps connection open' do
        subscriber.clear_events!

        expect do
          collection.insert_one(test: 1)
        end.to raise_error(Mongo::Error::OperationFailure, /10107/)

        expect(subscriber.select_published_events(Mongo::Monitoring::Event::Cmap::PoolCleared).count).to eq(0)

        expect do
          collection.insert_one(test: 1)
        end.to_not raise_error
      end
    end

    describe 'not master - 4.0' do
      max_server_version '4.0'

      let(:write_concern) { {:w => 1} }

      # not master
      let(:fail_point_code) { 10107 }

      it 'closes the connection' do
        subscriber.clear_events!

        expect do
          collection.insert_one(test: 1)
        end.to raise_error(Mongo::Error::OperationFailure, /10107/)

        expect(subscriber.select_published_events(Mongo::Monitoring::Event::Cmap::PoolCleared).count).to eq(1)

        expect do
          collection.insert_one(test: 1)
        end.to_not raise_error
      end
    end

    describe 'node shutting down' do
      let(:write_concern) { {:w => 1} }

      # interrupted at shutdown
      let(:fail_point_code) { 11600 }

      it 'closes the connection' do
        subscriber.clear_events!

        expect do
          collection.insert_one(test: 1)
        end.to raise_error(Mongo::Error::OperationFailure, /11600/)

        expect(subscriber.select_published_events(Mongo::Monitoring::Event::Cmap::PoolCleared).count).to eq(1)

        expect do
          collection.insert_one(test: 1)
        end.to_not raise_error
      end
    end
  end
end